Search Shows Content Type Machine Name instead of human readable name

This seems more a facetapi implementation bug.

So let's analyse this. In the apachesolr_search_process_response we do the following snippet and take the entity_type directly from the document.

      $result = array(
        // link is a required field, so handle it centrally.
        'link' => $path,
        // template_preprocess_search_result() runs check_plain() on the title
        // again.  Decode to correct the display.
        'title' => htmlspecialchars_decode($doc->label, ENT_QUOTES),
        // These values are not required by the search module but are provided
        // to give entity callbacks and themers more flexibility.
        'score' => $doc->score,
        'snippets' => $snippets,
        'snippet' => $snippet,
        'fields' => $fields,
        'entity_type' => $doc->entity_type,
        'bundle' => $bundle,
      );

And then in D6 we have the following function to process these vars :

/**
 * Process variables for search-result.tpl.php.
 *
 * The $variables array contains the following arguments:
 * - $result
 * - $type
 *
 * @see search-result.tpl.php
 */
function template_preprocess_search_result(&$variables) {
  $result = $variables['result'];
  $variables['url'] = check_url($result['link']);
  $variables['title'] = check_plain($result['title']);

  $info = array();
  if (!empty($result['type'])) {
    $info['type'] = check_plain($result['type']);
  }
  if (!empty($result['user'])) {
    $info['user'] = $result['user'];
  }
  if (!empty($result['date'])) {
    $info['date'] = format_date($result['date'], 'small');
  }
  if (isset($result['extra']) && is_array($result['extra'])) {
    $info = array_merge($info, $result['extra']);
  }
  // Check for existence. User search does not include snippets.
  $variables['snippet'] = isset($result['snippet']) ? $result['snippet'] : '';
  // Provide separated and grouped meta information..
  $variables['info_split'] = $info;
  $variables['info'] = implode(' - ', $info);
  // Provide alternate search result template.
  $variables['template_files'][] = 'search-result-'. $variables['type'];
}

The real responsible for this function is this. So we have two options now. Or you, as a site builder override this theme function and you load the respective type properly. Or we need to fix what we give to the theme function as the theme function itself cannot be changed (belongs to Drupal core)

if (!empty($result['type'])) {
    $info['type'] = check_plain($result['type']);
  }

Let's look at what 6.x-1.x did :

'type' => apachesolr_search_get_type($doc->type),

Aha! This looks like what we need.

The parsing code for this type happens in the result callback :

/**
 * Callback function for node search results.
 *
 * @param stdClass $doc
 *   The result document from Apache Solr.
 * @param array $result
 *   The result array for this record to which to add.
 */
function apachesolr_search_node_result($doc, &$result, &$extra) {
  $doc->uid = $doc->is_uid;
  $result += array(
    'type' => $doc->bundle,
    'user' => theme('username', $doc),
    'date' => isset($doc->changed) ? $doc->changed : 0,
    'node' => $doc,
    'uid' => $doc->is_uid,
  );

  if (isset($doc->is_comment_count)) {
    $extra['comments'] = format_plural($doc->is_comment_count, '1 comment', '@count comments');
  }
}

So let's add the same code of 6.x-1.x back and tada, we have a good solution.

Committed and marked as fixed. Thanks for the report

Closing to clean the issue queue a bit